It is common practice to inspect a workpiece after its production to determine the accuracy of the production process, that is, workpiece dimensions, correctness of angles, etc. For instance, such a measurement can be performed using a coordinate measuring machine.
For inspection, the workpiece is put on a base of such a coordinate measuring machine and a probe head being movable relative to the base is led to predetermined measurement points of the workpiece to obtain the exact coordinate data of these points. Thus, it is possible to determine the production accuracy of the workpiece.
In a conventional 3-D measurement machine, the probe head is supported for movement along three mutually perpendicular axes (in directions X, Y and Z). Thereby, the probe head can be guided to any arbitrary point within the working volume of the coordinate measuring machine.
In order to determine the coordinates, known measurement means capable to determine the probe head's distance from a known point of origin are employed. For instance, scales or other suitable measuring means are used for this purpose. The obtained coordinate data can then be stored in a memory such as a RAM and used for further processing.
With rather complicated structures of the workpiece, however, there arises a problem that it is cumbersome to guide the probe head to the selected target points. That is, it is required to timely reduce the movement speed of the probe head in order to prevent damage of the probe head or of the workpiece due to a too strong impact, when the probe head gets into abutment against the workpiece. In particular such a problem may arise with a fully automated coordinate measuring machine.
For precisely measuring and locating objects, various photogrammetric methods are known. Usually, two cameras are needed. Solutions with only a single camera also are known, wherein the camera needs to be operated at different locations in order to get the distance information. In addition, algorithms have to manage point clouds with a huge quantity of points and, as a consequence, need massive calculation power. For instance, the application WO 2013/083730 A1 discloses a CMM having at least one camera that can be directed to the object to be measured for locating the object.